Laser flash photolysis evidence for styryl radical cation cyclization in the SET-induced photorearrangement of a p-methoxy-substituted 2-phenylallyl phosphite

Article abstract of DOI:10.1021/jo0006775

The SET-induced photorearrangement of dimethyl 2-(4-methoxyphenyl)allyl phosphite, 9 (UV light, uranium glass filter, 9,10-dicyanoanthracene (DCA), biphenyl), gives phosphonate 12 in 83% isolated yield. Laser flash irradiation at 355 nm of oxygen saturate

Full text of DOI:10.1021/jo0006775

J . Org. Chem. 2000, 65, 6167-6172
6167
La ser F la sh P h otolysis Evid en ce for Styr yl Ra d ica l Ca tion
Cycliza tion in th e SET-In d u ced P h otor ea r r a n gem en t of a
p-Meth oxy-Su bstitu ted 2-P h en yla llyl P h osp h ite
Deepak Shukla,‡ Cuong Lu,† Norman P. Schepp,‡ Wesley G. Bentrude,*,† and
Linda J . J ohnston*,‡
Department of Chemistry, University of Utah, Salt Lake City, Utah 84112, and
National Research Council Canada, Ottawa, Canada K1A 0R6
bentrude@chemistry.utah.edu
Received May 2, 2000
The SET-induced photorearrangement of dimethyl 2-(4-methoxyphenyl)allyl phosphite, 9 (UV light,
uranium glass filter, 9,10-dicyanoanthracene (DCA), biphenyl), gives phosphonate 12 in 83% isolated
yield. Laser flash irradiation at 355 nm of oxygen saturated solutions of phosphite 9 containing
DCA and biphenyl generates the transient UV spectrum of the biphenyl radical cation that is
quenched by electron transfer from phosphite 9 (kq ) 8.9 × 109 M-1 s-1 at 20 °C) to form the
4-methoxystyryl cation 10. The UV spectrum of 10 decays by a measured first-order rate constant
of 8.0 × 106 s-1, presumably to generate the cyclic distonic radical cation 11. Intermediate 10 was
further characterized by measurement of the second-order rate constants for its reaction with azide,
chloride, and bromide ions and with the neutral nucleophile trimethyl phosphite. This study provides
the first spectroscopic evidence regarding the proposed mechanism (Schemes 1 and 2) for the SET-
induced photorearrangements of dimethyl 2-arylallyl phosphites to the corresponding 2-aryl-
allylphosphonates. Moreover, absolute rate constants for the intramolecular trapping of alkene
radical cations have seldom been measured. The removal of the electron from the styryl moiety of
phosphite 9, rather than from phosphorus, and the detectability of 10 arise from the stabilizing
effect of the 4-methoxy substituent. These results, however, do not allow conclusions to be made
concerning the site of removal of an electron in the SET-induced photorearrangement of dimethyl
2-phenylallyl phosphite 1 to phosphonate 6.
In tr od u ction
cations has been extensively studied, there are still
relatively few examples of intramolecular cyclizations.
Reported examples include trapping of alkene radical
cations by alcohols,4 carboxylic acids,5 and remote alk-
enes.6,7
The quantum yield for formation of phosphonate 6 (φP)
on SET-induced photorearrangement of 1 with DCA,
0.03, has been shown to be identical with that for
generation of phosphonate 8 from phosphite 7.1 This is
The rearrangement of dimethyl 2-phenylallyl phos-
phite 1 to the corresponding 2-phenylallyphosphonate,
6, can be photoinduced via an SET process with the
singlet excited state of 9,10-dicyanoanthracene (1DCA*)
as the electron acceptor (Scheme 1).1 The regiochemistry
of the process, determined previously by deuterium NMR
spectroscopy, is shown in Scheme 1. The initial site of
oxidation of 1 is unknown, because the oxidation half-
wave potentials for the model systems trimethyl phos-
phite, (MeO)3P (Eox ) 1.64 eV2), and R-methylstyrene,
MeCPhd CH2 (Eox ) 1.76 eV3), are quite similar. Nev-
ertheless, cyclization of the initial radical cation (2a or
2b) is proposed to generate the distonic 1,3-radical cation
3.1 Two potential routes for formation of 6 are given in
Scheme 1: (1) â scission of 3 followed by reduction of the
resulting 2-phenylallylphosphonate radical cation, 5, and
(2) reduction of radical cation 3 to the phosphoranyl 1,3-
diradical, 4, followed by â scission to introduce the two
π bonds in 6. Although intermolecular trapping of radical
in contrast to the very different quantum yields noted8
(5) (a) McManus, K. A.; Arnold, D. R. Can. J . Chem. 1995, 73, 2158.
(b) Herbetz, T.; Roth, H. D. J . Am. Chem. Soc. 1996, 118, 10954. (c)
Dai, S.; Wang, J . T.; Williams, F. J . Chem. Soc., Perkin Trans. 1 1989,
1063. (d) J iang, Z. Q.; Foote, C. S. Tetrahedron Lett. 1983, 24, 461.
(6) (a) Gassman, P. G.; Bottoroff, K. J . J . Am. Chem. Soc. 1987, 109,
7547. (b) Gassman, P. G.; DeSilva, S. A. J . Am. Chem. Soc. 1991, 113,
9870.
* Linda.J ohnston@nrc.ca.
(7) (a) Arnold, D. R.; McManus, K. A. Can. J . Chem. 1998, 76, 1238.
(b) Arnold, D. R.; Connor, D. A.; McManus, K. A.; Bakshi, P. K.;
Cameron, T. S. Can. J . Chem. 1996, 74, 602. (c) Weng, H.; Scarlata,
C.; Roth, H. D. J . Am. Chem. Soc. 1996, 118, 10947. (d) Kim, T.;
Mirafzal, G. A.; Liu, J .; Bauld, N. L. J . Am. Chem. Soc. 1993, 115,
7653.
(8) (a) Tamai, T.; Mizuno, K.; Hashida, I.; Otsuji, Y.; Ishida, A.;
Takamuku, S. Chem Lett. 1994, 149. (b) Mizuno, K.; Tamai, T.;
Hashida, I.; Otsuji, Y.; Kuriyama, Y.; Tokumaru, K. J . Org. Chem.
1994, 59, 7329. (c) Gollnick, K.; Schnatterer, S.; Utschick, G. J . Org.
Chem. 1993, 58, 6049.
† University of Utah.
‡ National Research Council, Canada.
(2) Ganapathy, S.; Dockery, K. P.; Sopchik, A. E.; Bentrude, W. G.
J . Am. Chem. Soc. 1993, 115, 8863. The stereochemistry at phosphorus
for this rearrangement has recently been reported: Hager, D. C.;
Sopchik, A. E.; Bentrude, W. G. J . Org. Chem. 2000, 65, 2778.
(3) Oxidation potential vs SCE for (MeO)3P, 1.64 eV (Ohmori, H.;
Nakai, S.; Masui, M. J . Chem. Soc., Perkin Trans. 1979, 2023).
(4) Oxidation potential for R-methylstryene vs SCE, 1.76 eV (Katz,
M.; Riemenschneider, P.; Wendt, H. Electrochim. Acta 1972, 17, 1595).
10.1021/jo0006775 CCC: $19.00 © 2000 American Chemical Society
Published on Web 08/24/2000

6168 J . Org. Chem., Vol. 65, No. 19, 2000
Shukla et al.
Sch em e 1
to give the cyclic distonic 1,3-radical cation 11. These
results provide strong corroborative evidence for the
presumed mechanism of the photorearrangement of 9.
The readily oxidized 4-methoxystyryl group ensures
donation of the electron from the styryl moiety of phos-
phite 9 to the excited singlet of DCA. The measured rate
constant for cyclization of the initial radical cation also
provides one of the first examples of absolute kinetic data
for intramolecular nucleophilic trapping of an alkene
radical cation.9
Resu lts a n d Discu ssion
Direct experimental evidence concerning the proposed
mechanism of Scheme 1 has not been presented up to
now. The Weller equation11 estimates oxidation of (MeO)3P
1
by DCA* to be energetically favorable by 10 kcal/mol12
while electron removal from R-methylstyrene should be
accompanied by an energy release of 7 kcal/mol.12 As
noted earlier1 these values are too similar to predict with
assurance that the site of removal of the electron by the
excited singlet DCA from 1 should be the phosphorus
atom rather than the styryl π bond.
In an initial laser flash photolytic approach to study
of the mechanism of Scheme 1, using dimethyl 2-phenyl-
allyl phosphite 1 and several photosensitizers, none of
the potential intermediates, 2-5, could be detected in
either acetonitrile or 2,2,2-trifluoroethanol. Failure to
detect the initial radical cation may indicate that its
lifetime is too short for detection due to rapid cyclization.
Alternately, the radical cation may not have a sufficiently
strong UV-visible absorption for detection under our
conditions, particularly if it is localized on phosphorus
(2a ). Although styrene radical cations have been char-
acterized by LFP, it is worth noting that the R-methyl-
styrene radical cation is too short-lived for nanosecond
measurements in acetonitrile but can be observed in
TFE.15 Thus, observation of a radical cation localized on
the styryl moiety of 1 can be difficult, even in the absence
of rapid intramolecular cyclization.
Sch em e 2
The stabilizing effect on styryl radical cations of a
4-methoxy substituent has been well documented.15,16 The
removal of an electron from p-methoxystyrene (Eox ) 1.49
1
eV16) by DCA* is 13 kcal/mol exergonic, i.e., 6 kcal/mol
more favorable than the -7 kcal/mol value estimated
earlier for R-methylstyrene as a model for 1. Further-
more, electron transfer from the styryl group of 9
(Scheme 2) is likely to be 2-3 kcal/mol more exergonic
than that from p-methoxystyrene by analogy with the
effects of R-methyl substitution (actually CH2O in 9) on
the oxidation potentials of related styrenes.16 Thus, the
for the regiospecific, triplet-sensitized photorearrange-
ments of 1 (φP ) 0.2-0.3) and 7 (φP ) 0.003) to 6 and 8,
respectively.
In the present paper we report the successful detection
of the 4-methoxystyryl radical cation 10 (Scheme 2),
generated from dimethyl 2-(4-methoxyphenyl)allyl phos-
phite 9 by laser flash photolysis (LFP) techniques and
the observation of the first-order decay of 10, presumably
(9) Several rate constants for intramolecular dimerization6d,7,10 and
Diels-Alder cyclizations6d,10 of alkene radical cations have been
reported.
(10) Schepp, N. P.; Shukla, D.; Sarker, H.; Bauld, N. L. J ohnston,
L. J . J . Am. Chem. Soc. 1997, 119, 10325.
(11) Rehm, D.; Weller, A. Isr. J . Chem. 1979, 8, 4269.
(12) Calculated using -0.89 eV vs SCE as the reduction potential
for DCA,13 1.64 eV for the oxidation potential of (MeO)3P,2 1.76 eV for
the oxidation potential of R-methylstyrene,3 0.06 eV as the Coulombic
term in acetonitrile,14 and 2.88 eV for the energy of the first singlet of
DCA (1DCA*).13
(13) Darmanyan, A. P. Chem. Phys. Lett. 1984, 110, 89.
(14) Table 2 of Photoinduced Electron Transfer; Fox, M. A., Chanon,
M., Eds.; Elsevier: New York, 1988; Part C, Chapter 4.1.
(15) J ohnston, L. J .; Schepp, N. P. J . Am. Chem. Soc. 1993, 115,
6564.
(8) (a) Bentrude, W. G.; Dockery, K. P.; Ganapathy, S.; Lee, S.-G.;
Tabet, M.; Wu, Y. W.; Cambron, R. T.; Harris, J . M. J . Am. Chem.
Soc. 1996, 118, 6192. (b) Ganapathy, S.; Cambron, R. T.; Dockery, K.
P.; Wu, Y. W. Tetrahedron. Lett. 1993, 34, 5987. (c) Bentrude, W. G.;
Lee, S.-G. J . Am. Chem. Soc. 1987, 109, 1577. (d) Bentrude, W. G.;
Lee, S.-G.; Akutagawa, K.; Ye, W.; Charbonnel, Y.; Omelanczuk, J .
Phosphorus Sulfur 1987, 30, 105. The stereochemistry at phosphorus
of this rearrangement has recently been reported: Hager, D. C.;
Bentrude, W. G. J . Org. Chem. 2000, 65, 2786.
(16) Schepp, N. P.; J ohnston, L. J . J . Am. Chem. Soc. 1996, 118,
2872.

LFP Evidence for Styryl Radical Cation Cyclization
J . Org. Chem., Vol. 65, No. 19, 2000 6169
4-methoxy-substituted analogue of 1 should provide a
better substrate for obtaining spectroscopic and kinetic
information on the initial radical cation and its further
rearrangements. Therefore, dimethyl 2-(4-methoxyphen-
yl)allyl phosphite 9 was prepared by standard techniques
from reaction of dimethyl N,N-dimethylphosphoramidite
with the requisite alcohol, 2-(4-methoxyphenyl)-2-propen-
1-ol. Purified phosphite was characterized spectroscopi-
cally and by conversion to the phosphate, 9-O, easily
identified by NMR, elemental analysis, and HRMS
(Experimental Section).
P r od u ct Stu d y. Authentic phosphonate 12 was pre-
pared near-quantitatively (93% isolated yield) by the
sensitization of the photorearrangement of 9 by ben-
zophenone triplet, populated by Pyrex-filtered light from
a 450 W medium-pressure mercury UV lamp (see Ex-
perimental Section). The near-quantitative triplet-sen-
sitized conversion of dimethyl 2-phenylallyl phosphite (1)
to the corresponding 2-phenylallylphosphonate (6) was
reported earlier.8 The structure of 12 was readily verified
F igu r e 1. Transient absorption spectra of 10 obtained by 355
nm excitation of 9,10-dicyanoanthracene (3 × 10-4 M) in the
presence of 0.15 M biphenyl and 1.5 mM dimethyl 2-(4-
methoxyphenyl)allyl phosphite 9 in oxygen-purged acetonitrile.
Spectra were recorded 0.24 (b), 1.04 (9), 3.0 (]), and 6.4 (2)
µs after the laser pulse.
1
by mass spectrometry (LRMS and HRMS) along with H,
13C, and 31P NMR spectroscopy.
The excited singlet DCA-induced photorearrangement
of a deoxygenated 0.14 M acetonitrile solution of 9 that
was also 0.02 M in DCA and contained a large excess of
biphenyl (4 M) as cosensitizer was carried out by irradia-
tion through Pyrex with 350 nm light from a Rayonet
reactor. The reaction was monitored by GC which showed
the consumption of phosphite 9 to be complete in 72 h.
Phosphonate 12 was obtained by radial chromatography
in 83% isolated yield.
to increase the efficiency of production of free radical ions
in other photoinduced SET reactions is well-docu-
mented.18
Intermediate 10 decayed in the presence of added
nucleophiles, Nu, in CH3CN with an observed rate
constant, kobs, correlated with nucleophile concentration
by the rate equation:
La ser F la sh P h otolysis (LF P ). An important aspect
of the mechanism of the DCA/biphenyl cosensitized
rearrangement of 9 was revealed in a LFP study. In
particular the removal of an electron from 9 gives the
detectable radical cation 10 with the radical cation
localized in the 2-(4-methoxyphenyl)allyl functionality.
This means that, as anticipated, the methoxy substitu-
tion has rendered the alkene site the one of definitively
lower oxidation potential and also has sufficiently sta-
bilized the radical cation to make it observable.
Thus, laser excitation (355 nm) of an oxygen-saturated
kobs ) k0 + kNu[Nu]
Nicely linear plots (correlation coefficients greater than
0.99) of [Nu] vs kobs for azide, chloride, and bromide ion,
that included the experimentally determined origin value
of k0 ([Nu] ) 0) mentioned earlier, yielded second-order
rate constants for nucleophilic attack: kazide ) 1.9 × 1010
M-1 s-1; kCl ) 2.2 × 1010 M-1 s-1; kBr ) 1.8 × 1010 M-1
s-1. The similarity of these rate constants to those
previously determined for attack of this series of nucleo-
philes on 4-methoxystyrene radical cation itself is con-
sistent with assignment of the observed transient to
radical cation 10. To further confirm that the rate
constant for decay of 10 is limited by intramolecular
cyclization, we examined the reaction of the radical cation
from p-methoxystyrene itself with trimethyl phosphite
in oxygen-saturated acetonitrile. The measured second-
order rate constant of 4.4 × 108 M-1 s-1 is consistent with
previous measurements of the reactivity of other un-
charged nucleophiles such as alcohols with styrene
radical cations15 and clearly indicates that phosphites are
efficient radical cation traps.
acetonitrile solution of DCA (5.0 × 10-4 M, OD355nm
)
0.52) containing biphenyl (0.2 M) at 20 °C gave the
characteristic transient spectrum17of biphenyl radical
cation (λmax ) 390 and 680 nm, τ ≈ 1.7 µs). The decay of
the biphenyl radical cation was enhanced in the presence
of 9. The slope of a linear plot of the observed rate
constant for decay of the biphenyl radical cation, kobs, vs
[9] ([9] ) 0.0002-0.0009 M, seven points, correlation
coefficient ) 0.99734) yielded a second-order rate con-
stant, kq ) 8.9 × 109 M-1 s-1. At [9] ) 0.002 M, the
biphenyl radical cation spectrum was completely quenched
with the concomitant formation of a new transient with
two absorption bands (Figure 1, λ ) 610 and 380 nm).
The identical decay kinetics for the two bands (k0 ) 8.0
× 106 s-1) showed that they belong to a single species.
The close similarity of this spectrum to that of the
4-methoxystyrene radical cation in CH3CN15 assigns it
to the radical cation of 9 with the ion radical localized
on the p-methoxystyrene moiety (10). An experiment in
the absence of biphenyl cosensitizer failed to give detect-
able concentrations of transient 10. The use of biphenyl
The observed first-order decay of 10 in the absence of
added nucleophiles, k0, presumably is for the cyclization
step 10 f 11. By comparison, the radical cation from
p-methoxystyrene is approximately 2 orders of magnitude
longer lived in the same solvent in the absence of
(18) The use of 1DCA* to generate biphenyl radical cation leads to
a quantum yield of cage escape to give free radical ions of 0.83. The
high yield is due to the fact that back electron transfer between
biphenyl radical cation and cyanoaromatic sensitizer radical anions
is sufficiently exergonic to fall in the Marcus inverted region and is
thus much slower than is typical for more easily oxidized donors. Gould,
I. R.; Ege, D.; Mattes, S. L.; Farid, S. J . Am. Chem. Soc. 1987, 109,
3794.
(17) Shida, T. Electronic Absorption Spectra of Radical Ions; Elsevi-
er: Amsterdam, 1988.

6170 J . Org. Chem., Vol. 65, No. 19, 2000
Shukla et al.
nucleophiles.15 The additional substituent in 10, the
â-CH2OP, would be expected to further lengthen the
lifetime of the radical cation.
analogue of 5. However, the latter radical cation may be
rapidly reduced by DCA-•. This process should be ∼2.3
eV exergonic, based on an estimated oxidation potential
of 1.4 eV for p-methoxy-R-methylstyrene16 and the known
reduction potential of DCA (-0.89 eV13).
The rate constant for cyclization of radical cation 10
can be compared to literature data for related radical
cation cyclizations of alkenes. For example, rate con-
stants that vary from 3 × 106 to 1.2 × 109 s-1 have been
reported for intramolecular dimerization and cycloaddi-
tion of p-methoxy-â-alkylstyrene radical cations.6d,10 Simi-
lar data are available for radical cation-initiated dimer-
ization of several 1,ω-bis(diarylethenyl)alkanes.7 Although
absolute kinetic data are not available for the intramo-
lecular nucleophilic trapping of alkene radical cations,
the competition between inter- and intramolecular alco-
hol trapping has given rise to an estimate of ∼2 × 106
s-1 for the rate of intramolecular trapping of a vinylcy-
clopropane radical cation.4b The relatively slow reaction
has been suggested to reflect steric hindrance to intramo-
lecular capture.
A model for the reduction of 3 to 1,3-biradical 4 by
DCA-• might be the reduction of [RPPh3]+ for which
reduction potentials have been published.21 Conversion
of 3 to 4 by DCA-• is thereby estimated to be endergonic
by 11 kcal/mol. It is likely that the reduction potential
for 3 is in fact more favorable since [(RO)3PR′]+, a closer
model for 3, should have a less negative one-electron
reduction potential as a result of the electronegative
alkoxy substitutents on phosphorus. Furthermore, the
formation of 1,3-biradical 4 may be energetically more
favorable than the above model estimates since certain
cyclic five-membered ring 1,3-diradicals in all-carbon ring
systems are known to be stabilized relative to their
monoradical counterparts.22 Moreover, energetically un-
favorable reductions can occur as essential steps in
electron-transfer chemistry. The energetics of the poten-
tial alternative competing step 3 f 5 of Scheme 1 is not
readily approximated, although the introduction of two
π bonds should be energetically favorable.
Evidence against the process 3 f 5 comes from
unsuccessful attempts to trap 5 with MeOH (unpub-
lished19) in studies of the SET-induced rearrangement
of 1 with DCA as electron acceptor. Although an ap-
proximate 8% yield of methanol adduct 13 results when
1 is rearranged under SET conditions in the presence of
MeOH, its generation lags considerably behind formation
of 6. We assign the origin of 13 to the DCA-sensitized
In contrast to the results described for phosphite 9, use
of biphenyl cosensitizer with DCA under conditions
similar to those employed for 9 failed to cleanly effect
the rearrangement of 1 (unpublished results19), although
in the absence of biphenyl reasonably high yields of 6
were obtained. A second product, (MeO)2P(O)C6H4-C6H5
(two isomers), from competing phosphorylation of bi-
phenyl was observed along with 6. This product most
likely arises from nucleophilic attack by 1 on the biphenyl
radical cation. (The same phosphorylation product is
formed exclusively when (MeO)3P instead of 1 is used.19)
This indicates that electron transfer from 1 to the
biphenyl radical cation is energetically relatively unfa-
vorable, and reaction with phosphorus, acting as a
nucleophile, competes with electron transfer. (Literature
data on reactions of nucleophiles with styrene and
anthracene radical cations indicate that electron transfer
is favored over nucleophilic addition when the former is
exergonic by 0.1-0.2 eV.20) Clearly, the energetics of
electron transfer to biphenyl radical cation from the
4-methoxystyryl moiety of 9 is greatly favored over
donation of an electron from either the styryl functional-
ity or phosphorus atom of 1 or 9 and is also more rapid
than nucleophilic attack by phosphorus. However, these
results do not allow the site of removal of the electron
addition of MeOH to 6, a reaction which we found to be
readily induced in an independent experiment.19 The lack
of formation of trimethyl phosphate in the reactions with
methanol as a trap argues for the initial formation of 2b
rather than 2a .
Con clu sion s
1
from 1 by DCA* to be assigned.
A laser flash photolysis study of the 1DCA*-induced
photorearrangement of the allyl phosphite 9 to phospho-
nate 12 gives UV absorption spectroscopic evidence for
the formation of the transient radical cation 10. Inter-
mediate 10 is further characterized by reasonable values
of second-order rate constants for its reaction with a
series of added nucleophiles. The lifetime of 10 is ap-
proximately 100 times shorter than that of its simpler
counterpart, p-methoxystyrene itself. The measured first-
order rate constant for its disappearance (8.0 × 106 M-1
s-1) presumably corresponds to the cyclization step 10
f 11 and provides one of the first calibrations of an
intramolecular trapping rate for an alkene radical cation.
Such data are of key importance in development of a firm
mechanistic basis for the use of radical ion chemistry in
Unfortunately, the cyclic distonic radical cation 11,
which should display UV absorption characteristic of a
benzylic radical, is not detected by the LFP experiments,
perhaps because it undergoes very rapid reaction. The
detection of a benzylic radical is also complicated by the
fact that there is significant ground-state absorption at
320 nm under the conditions required for the photosen-
sitized electron-transfer experiments using DCA/biphen-
yl. Furthermore, the present study gives no definitive
information for phosphite 1 concerning the two pathways
for conversion of 3 to 6 in Scheme 1. However, if 3 f 5
were the operative mechanism for this photorearrange-
ment (Scheme 1), evidence for formation of the 4-methoxy
analogue of 5 might have been encountered in the LFP
kinetic study of 9. The kinetic measurements then should
have been poorly behaved because of the expected
similarities of the UV spectra of 10 and the 4-methoxy
(21) On the basis of the average of one-electron reduction potentials
vs SCE of [PhCMe2PPh3]+ (-1.5 eV) and [MeCPh2PPh3] (-1.24 eV)
(Saveant, J . M.; Binh, S. K. J . Org. Chem. 1977, 42, 1242) and the
reduction potential of DCA (-0.89 eV13).
(22) Adam. W.; Hannemann, K.; Wilson, R. M. J . Am. Chem. Soc.
1986, 108, 929.
(19) Dockery, K. P. Ph.D. Dissertation, University of Utah, 1995.
(20) J ohnston, L. J .; Schepp, N. P. Adv. Electron-Transfer Chem.
1996, 5, 41.

LFP Evidence for Styryl Radical Cation Cyclization
J . Org. Chem., Vol. 65, No. 19, 2000 6171
by stirring overnight, then filtration and distillation. 31P NMR
synthesis and in probing biological reaction mechanisms.
The results also provide supporting evidence for the
reaction mechanism proposed in Scheme 1 with initial
formation of radical cation 10, analogous to 2b, at least
when a stabilizing 4-methoxy substituent is attached to
the phenyl ring. Unfortunately, no spectroscopic evidence
for the intermediates analogous to 3-5 was encoun-
tered.23
(121.4 MHz, CDCl3): δ 150.21 [lit.25 δ 151.0].
P r ep a r a tion of 2-(4-Meth oxyp h en yl)-2-p r op en e. A pro-
cedure adapted from Maryanoff and Reitz26 was used. To a
stirred solution of methyl triphenylphosphonium bromide (48
g, 0.13 mol) in dry tetrahydrofuran (500 mL), cooled to -20
°C, was added 2.5 M of n-BuLi in hexane (53 mL, 0.13 mol)
via syringe. After 30 min of stirring at -20 °C, 4-methoxy-
acetophenone (10 g, 67 mmol) was added in one portion. The
reaction was quenched with water (100 mL) and extracted with
tetrahydrofuran (3 × 50 mL) which was washed with aqueous
sodium chloride (3 × 50 mL). The organic layer was dried over
magnesium sulfate, filtered, and concentrated in vacuo to give
a brown oil. Distillation in vacuo (bp 51 °C at 0.35 Torr) yielded
white crystals (8.1 g, 82% yield); mp 33.5-34.5 °C (lit.27 32.0-
Exp er im en ta l Section
Gen er a l Meth od s. Starting materials were obtained from
Aldrich or Acros and, unless otherwise indicated, were used
without further purification. Anhydrous solvents were ob-
tained as follows: diethyl ether (Et2O), tetrahydrofuran (THF),
and benzene were distilled over sodium/benzophenone under
nitrogen; methylene chloride and acetonitrile were distilled
over calcium hydride under argon. Triethylamine and diiso-
propylamine were distilled over calcium hydride prior to use.
Benzophenone was recrystallized from ethanol. All atmosphere-
sensitive reaction solutions were prepared under argon using
glovebag techniques and then run under argon. Analytical
thin-layer chromatography (TLC) was performed on Merck
silica gel 60 F254 precoated silica gel plates. Column chroma-
tography was carried out with silica gel 60 (230-400 mesh)
from Mallinckrodt. Chromatography performed on a Chro-
matotron7 7924T apparatus utilized silica gel 60 PF254 contain-
ing gypsum purchased from EM Science. Distillations were
performed on a short-path distillation apparatus. Melting
points are uncorrected. Elemental analyses were carried out
by Atlantic Microlab, Inc., Norcross, GA, and by Midwest
Microlab, LLC, Indianapolis, IN.
1
4
32.5 °C). H NMR (299.7 MHz, CDCl3): δ 2.13 (dd, J ) 1.5
4
4
2
Hz, J ) 0.7 Hz, 3H), 3.81 (s, 3H), 4.99 (dq, J ) 1.5 Hz, J )
2.4 Hz, 1H), 5.28 (dq, 2J ) 2.4 Hz, 4J ) 0.7 Hz, 1H), 6.85-
7.43 (m, 4H). 13C NMR (75.5 MHz, CDCl3): δ 21.91 (CH3),
55.25, 110.64, 113.51, 126.58, 133.70, 159.04, 142.52; LRMS
(EI) m/z 148 (M+, 53), 135 (78), 105 (44), 89 (65), 77 (87), 61
(20), 51 (20), 43 (47), 28 (100). HRMS (EI): calcd for C10H12
O
[M+] 148.0888. Found: 148.0909. Anal. Calcd for C10H12O: C,
81.04; H, 8.16. Found: C, 81.22; H, 8.29.
P r ep a r a tion of 4-Meth oxya cetop h en on e Tosylh yd r a -
zon e. A solution of 4-methoxyacetophenone (60 g, 0.40 mol)
in methanol (500 mL) was magnetically stirred at room
temperature for 5 min. Tosylhydrazide (93 g, 0.50 mol) was
added in one portion followed by 2 drops of concentrated
hydrochloric acid. The reaction solution was stirred for another
3 h at room temperature, monitored by TLC, and then placed
in
a freezer overnight. The hydrazone precipitated from
solution as a white solid that was washed with 250 mL of cold
methanol and then dried in vacuo overnight to give 125 g (98%
yield): mp 164.5-165.5 °C (dec) [lit.28 mp 166-168 °C (dec)].
HRMS (EI): calcd for C16H18N2O3S [M+] 318.1038. Found
318.1037. Anal. Calcd for C16H18N2O3S: C, 60.36; H, 5.70.
Found: C, 60.44; H, 5.75.
NMR a n d GC Mea su r em en ts. Proton (1H), carbon (13C),
and phosphorus (31P) NMR spectra were recorded routinely
and referenced to the usual standards. NMR splitting patterns
are designated as s (singlet), d (doublet), t (triplet), q (quartet),
quin (quintet), sept (septet), and m (multiplet). Coupling
constants (J ) are in hertz. J values refer to proton-proton
coupling unless otherwise stated. 13C NMR J values designate
carbon-phosphorus couplings. Gas chromatography (GC) was
performed in the FID mode on 20 m × 0.25 mm × 0.25 µm
fused silica DB-1 phase capillary column. GC-MS (EI) mea-
surements (70 eV) utilized a 30 m × 0.25 mm × 0.25 µm fused
silica HP-1 phase capillary column. High-resolution EIMS
measurements also were at 70 eV. The molecular ion is
indicated by [M+].
La ser F la sh P h otolysis. The flash photolysis system has
been described in detail elsewhere.24 For these experiments a
Lumonics HY750 Nd:YAG laser (355 nm, 10 ns/pulse, <50 mJ /
pulse) was used for sample excitation. Samples were contained
in 7 × 7 mm2 quartz cells and were deaerated by purging with
oxygen prior to laser irradiation.
P r epar ation of 2-(4-Meth oxyph en yl)-2-pr open -1-ol. The
method was adapted from that of Chamberlin and co-work-
ers.29 To a stirred solution of 4-methoxyacetophenone tosyl-
hydrazone (10 g, 31 mmol) and tetrahydrofuran (300 mL) at 0
°C under argon was added, dropwise, 2.5 M n-BuLi in hexane
(42 mL, 0.10 mol) via syringe. The resulting red solution was
stirred at 0 °C until the evolution of N2 gas was complete.
Paraformaldehyde (1.9 g, 63 mmol), dried overnight over
phosphorus pentoxide in a desiccator, was added in one
portion. The reaction mixture changed color from dark red to
creamy yellow. The stirred solution was warmed to room
temperature and held overnight or until completion of the
reaction (monitored by TLC). The reaction was quenched with
aqueous ammonium chloride (100 mL) and extracted with
ethyl acetate (3 × 50 mL) which was dried over anhydrous
magnesium sulfate. The solution was concentrated in vacuo
to give a dark brown oil that was purified by column chroma-
tography on silica gel (eluted with 10% triethylamine, 30%
ethyl acetate-hexanes) to give a solid, recrystallized from ethyl
acetate/hexanes to yield white crystals (2.9 g, 55% yield): mp
Dim eth yl N,N-d iisop r op ylp h osp h or a m id ite was rou-
tinely prepared by coaddition under argon of a solution of
(isoPr2N)2PCl (94 g, 0.46 mol) in diethyl ether (100 mL) and
triethylamine (94 g, 0.93 mol) in methanol (30 g, 0.93 mol) to
650 mL of stirred diethyl ether at room temperature, followed
1
78-79 °C [lit.29 mp 71-73 °C]. H NMR (300.1 MHz, CDCl3):
4
δ 1.80 (s, 1H), 3.81 (s, 3H), 4.50 (s, 2H), 5.25 (dt, J ) 1.5 Hz,
(23) A reviewer has suggested that these SET processes involve back
electron transfer from the DCA anion radical to 2b or 10 to generate
the triplet excited state of 1 and 9, respectively, that then leads to
phosphonate via mechanisms previously proposed by the Bentrude
group8 for triplet excited states of allyl phosphites formed by triplet
sensitization. This is highly unlikely, however, because of (1) the
predicted low efficiency of initial triplet ion pair generation, (2) the
likely inability of the bimolecular electron transfer to 2b or 10 to
compete with cyclization to 3 or 11, and (3) the likelihood that electron
transfer from DCA anion radical to phosphite radical cation would not
be exoergic enough to generate triplet phosphite efficiently. Further-
more, the close similarity in values (ca. 0.025) for the quantum
efficiency of phosphonate formation from 1 and from its counterpart
with phosphorus in a six-membered ring when excited singlet DCA is
the SET agent1 is contrary to the known at least 100-fold greater
efficiency of triplet sensitized reaction of 1.8
2J ) 1.2 Hz, 1H), 5.38 (dt, 2J ) 1.2 Hz, 4J ) 0.7 Hz, 1H), 6.85-
7.42 (m, 4H). 13C NMR (75.5 MHz, CDCl3): δ 55.27, 65.10,
111.05, 113.85, 127.18, 130.83, 159.37, 146.50. Anal. Calcd for
C
10H12O2: C, 73.15; H, 7.37. Found: C, 72.87; H, 7.37.
P r ep a r a tion of Dim eth yl 2-(4-Meth oxyp h en yl)-2-p r o-
p en -1-yl P h osp h ite (9). A solution of 2-(4-methoxyphenyl)-
(25) Mag, M.; Engels, J . W. Nucleosides Nucleotides 1988, 7, 725.
(26) Maryanoff, B. E.; Reitz, A. B. Chem. Rev. 1989, 89, 863.
(27) Deno, N. C.; Frederick, A. K.; Peterson, H. J . J . Am. Chem.
Soc. 1965, 87, 2157.
(28) McMahon, R. J .; Chapman, O. L. J . Am. Chem. Soc. 1987, 109,
683.
(24) Kazanis, S.; Azarani, A.; J ohnston, L. J . J . Phys. Chem. 1991,
95, 4430.
(29) Chamberlin, A. R.; Stemke, J . E.; Bond, F. T. Tetrahedron Lett.
1986, 27, 5579.

6172 J . Org. Chem., Vol. 65, No. 19, 2000
Shukla et al.
2-propen-1-ol (1.0 g, 6.1 mmol), dimethyl N,N-diisopropylphos-
phoramidite (1.4 g, 7.3 mmol), dry acetonitrile (15 mL), and
1H-tetrazole (61 mg, 0.87 mmol) was stirred under argon at
ambient temperature overnight and filtered using Schlenck
techniques to remove the amine salts which were washed with
cold diethyl ether (3 × 10 mL). Solvent was removed using a
vacuum pump and a coldfinger solvent trap to yield an oily
residue. Distillation (bp 105 °C/1.4 Torr) produced a clear oil
(1.3 g, 83% yield, 99% GC purity). 1H NMR (300.1 MHz,
CDCl3): δ 3.50 (d, 3J HP ) 10.5 Hz, 6H), 3.82 (s, 3H), 4.69 (ddd,
3J HP ) 7.8 Hz, 4J ) 1.2 Hz, 4J ) 0.7 Hz, 2H), 5.32 (dt, 4J ) 1.2
(21), 115 (24), 109 (28), 103 (22), 94 (23), 91 (20). HRMS (EI):
calcd for C12H17O4P [M+] 256.0864. Found: 256.0856.
Meth od B. SET In d u ced . Similarly, a 5.0 mL acetonitrile
solution of dimethyl 2-(4-methoxyphenyl)-2-propen-1-yl phos-
phite, 9 (0.17 g, 0.76 mmol), 9,10-dicyanoanthracene (11 mg,
0.05 mmol), and biphenyl (3.1 g, 0.02 mol) was prepared under
argon. The solution was added to an argon-flushed Pyrex test
tube, capped as in Method A, and irradiated with light from a
Rayonet photoreactor (six 350 nm lamps). Upon total con-
sumption of phosphite (73 h), the solvent was removed in vacuo
to give a yellow oil. Purification by radial chromatography on
silica gel (Chromatotron,7 eluted with 30% ethyl acetate-
hexanes) gave a colorless oil (0.14 g, 83% yield, 99% GC purity)
that was spectroscopically identical to the material prepared
by Method A.
2
4
Hz, J ) 1.2 Hz, 1H), 5.43 (d, J ) 0.7 Hz, 1H), 6.86-7.42 (m,
2
4H). 13C NMR (75.5 MHz, CDCl3): δ 49.28 (d, J ) 10.3 Hz),
2
55.27, 63.91 (d, J ) 12.3 Hz), 112.41, 113.73, 127.21, 130.76,
3
159.37, 143.80 (d, J ) 4.8 Hz). NMR (121.5 MHz, CDCl3): δ
141.25. Elemental analysis was obtained on the phosphate,
9-O.
P r ep a r a tion of Dim eth yl 2-(4-Meth oxyp h en yl)-2-p r o-
p en -1-yl P h osp h a te (9-O). To a stirred solution of phosphite
9 (0.14 g, 0.55 mmol) in pentane (10 mL) were added two drops
of 3.0 M anhydrous tert-butyl hydroperoxide in 2,2,4-trimeth-
ylpentane. Total consumption of 9 was confirmed by TLC. The
solution was dried with anhydrous magnesium sulfate and
filtered. Solvent removal in vacuo and purification by radial
chromatography on silica gel (Chromatotron,7 eluted with 30%
ethyl acetate-hexanes) gave a colorless oil (0.14 g, 95% yield).
P h otor ea r r a n gem en t of P h osp h ite 9 to Dim eth yl 2-(4-
Meth oxyp h en yl)-2-p r op en -1-ylp h osp h on a te 12. Meth od
A. Tr ip let Sen sitiza tion . Using glovebag techniques under
an argon atmosphere, a 5 mL solution in benzene of dimethyl
2-(4-methoxyphenyl)-2-propen-1-yl phosphite, 9 (0.58 g, 2.3
mmol), and benzophenone (35 mg, 0.19 mmol) was prepared
in a Pyrex test tube that was capped with a rubber septum
which was tied with a copper wire and was further sealed with
five layers of Parafilm. The solution was purged with argon
for 15 min and irradiated with Pyrex-filtered light from a
medium pressure 450 W Hg lamp. The reaction was followed
by GC and also by 31P NMR measurements on aliquots
withdrawn from the reaction solution. On total consumption
of phosphite (7 h), the solvent was removed in vacuo to give 9
as an oil that was purified by radial chromatography (Chro-
matotron,7 eluted with 30% ethyl acetate-hexanes) to give a
colorless oil (0.54 g, 93% yield, 99% GC purity). 1H NMR (300.1
3
1H NMR (299.7 MHz, CDCl3): δ 3.72 (d, J HP ) 11.0 Hz, 6H),
3
4
2
3.82 (s, 3H), 4.92 (dd, J HP ) 7.6 Hz, J ) 0.7 Hz), 5.35 (dt, J
4
4
) 1.2 Hz, J ) 0.7 Hz, 1H), 5.50 (d, J ) 0.7 Hz, 1H), 6.87-
7.42 (m, 4H). 13C NMR (75.4 MHz, CDCl3): δ 54.31 (d, 2J )
2
6.1 Hz), 55.29, 69.02 (d, J ) 5.5 Hz), 113.87, 127.21, 129.88,
3
31
159.59 (aromatic), 113.97, 141.95 (d, J ) 7.3 Hz). P NMR
(121.3 MHz, CDCl3): δ 1.83. LRMS (EI): m/z 272 (M+, 100),
146 (20), 131 (14), 103 (12). HRMS (EI): calcd for C12H17O5P
[M+] 272.0814. Found: 272.0798. Anal. Calcd for C12H17O5P:
C, 52.94; H, 6.29. Found: C, 52.28; H, 6.41.
2
4
MHz, CDCl3): δ 3.06 (dd, J HP ) 22.1 Hz, J ) 0.7 Hz), 3.65
3
2
4
(d, J HP ) 10.7 Hz, 6H), 3.82, 5.26 (dd, J ) 5.6 Hz, J ) 1.0
Hz, 1H), 5.47 (dd, 2J ) 5.6 Hz, 4J ) 0.7 Hz, 1H), 6.86-7.15
Ack n ow led gm en t. We gratefully acknowledge sup-
port of this research by grants from the National Science
Foundation (W.G.B.) and the Public Health Service
(W.G.B).
(m, 4H). 13C NMR (75.5 MHz, CDCl3): δ 32.35 (d, J ) 138.7
1
2
4
Hz), 52.78 (d, J ) 6.6 Hz), 55.28, 113.69, 127.37 (d, J ) 1.0
3
3
Hz), 132.89 (d, J ) 4.0 Hz), 159.36, 115.68 (d, J ) 11.1 Hz),
137.64 (d, 2J ) 10.3 Hz). 31P NMR (121.5 MHz, CDCl3): δ
29.95. LRMS (EI): m/z 256 (M+, 100), 160 (49), 148 (83), 133
J O0006775